Electric taxi battery swap station optimization scheduling strategy based on comprehensive SOC and multi-time scale

Abstract

The invention discloses a comprehensive SOC and multi-time scale-based optimal scheduling method for an electric taxi battery swap station. The method comprises the following steps of establishing a battery charging and discharging model, a photovoltaic power generation model, a power grid interaction model and a battery swap service model for a battery swap station system; establishing a comprehensive SOC model based on the battery state, and proposing a charging scheduling strategy and a dynamic service charge strategy based on the comprehensive SOC; a battery swap station joint optimization scheduling model integrating SOC optimization management and multi-time scale scheduling efficiency is adopted; performing multi-target expansion on the original crayfish optimization algorithm, and introducing a self-adaptive improvement mechanism to form a self-adaptive multi-target crayfish optimization algorithm; and solving the model by using an improved algorithm. The solved battery swap station scheduling strategy can ensure that the average waiting time is minimum when the annual net profit is maximum, so that the overall comprehensive benefit is improved.

Description

Technical Field

[0001] This invention relates to an optimized scheduling strategy for electric taxi battery swapping stations based on integrated SOC and multiple time scales. Through dynamic modeling and multi-time coordination optimization, it aims to achieve optimal operation of battery swapping stations under multiple objectives such as economy, efficiency and user satisfaction. Background Technology

[0002] With the increasing popularity of electric vehicles, battery swapping stations, as crucial energy supply infrastructure, directly impact the promotion and application of electric vehicles through their operational efficiency and economic viability. Battery swapping stations typically include battery storage systems, charging equipment, battery swapping equipment, and renewable energy generation systems. Their operation requires comprehensive consideration of multiple factors, including battery status, electricity demand, electricity price fluctuations, and renewable energy output. In actual operation, battery swapping stations not only need to meet users' demands for fast and convenient battery swapping but also address complex challenges such as grid load fluctuations, time-of-use electricity price differences, and the intermittency of renewable energy sources. Furthermore, the battery, as the core asset of a battery swapping station, directly affects the long-term economic viability and reliability of the system through its lifespan, health status, and charging / discharging strategies. Therefore, constructing an intelligent dispatching system capable of collaboratively optimizing battery management, power dispatching, and user services has become key to improving the operational level of battery swapping stations. By introducing multi-timescale dispatching strategies and a comprehensive SOC dynamic response mechanism, the economic viability, efficiency, and sustainability of system operation can be effectively balanced, thereby promoting the high-quality development of electric vehicle infrastructure.

[0003] Existing battery swapping station scheduling methods often focus on optimization at a single time scale or for a single objective, failing to fully integrate the dynamic characteristics of battery SOC (State of Charge) with multi-time-scale scheduling strategies. This makes it difficult to achieve global optimization in complex and ever-changing operating environments. Therefore, researching an optimized scheduling method for battery swapping stations that comprehensively considers battery SOC status, multi-time-scale scheduling, and dynamic electricity price response is of significant practical importance. This method can comprehensively consider multiple objectives such as the overall operating efficiency and economics of the swapping station, as well as the average waiting time for users, achieving optimal operation of the swapping station. Summary of the Invention

[0004] The present invention aims to provide an optimized scheduling strategy for electric taxi battery swapping stations based on integrated SOC and multiple time scales, in order to solve the problems of insufficient consideration of the dynamic characteristics of battery status, single scheduling strategy, and difficulty in coping with multi-objective optimization in the prior art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] This invention discloses an optimized scheduling strategy for electric taxi battery swapping stations based on integrated SOC and multiple time scales, comprising the following steps:

[0007] S1: Modeling the battery charging and discharging model, photovoltaic power generation model, power grid interaction model, and battery swapping service model for the battery swapping station system;

[0008] S2: Establish a comprehensive SOC model based on the battery status of the battery swapping station, and propose a charging scheduling strategy and a dynamic service fee strategy based on the comprehensive SOC status.

[0009] S3: Construct a joint optimization scheduling model for battery swapping stations that takes into account both comprehensive SOC optimization management and multi-time-scale scheduling efficiency;

[0010] S4: A self-improved Adaptive Multi-Objective Crayfish Optimization (AMOCO) algorithm is used for optimized scheduling of battery swapping stations. Based on Pareto theory, the original crayfish optimization algorithm is extended to a multi-objective model, resulting in a basic framework. In the initialization phase, a Logistic chaotic mapping method is introduced to perform chaotic inverse initialization of the population, and parameters are adaptively improved during the heat avoidance search process.

[0011] Specifically, in S1, the battery swapping station system modeling includes battery charging model, photovoltaic power generation model, power grid interaction model, and battery swapping service model modeling, and each model is established as follows.

[0012] Charging rate equation:

[0013]

[0014] In the formula, μ is the charging rate, and E c This refers to the state of charge of the battery during the charging process.

[0015] Real-time charging power of a single charger:

[0016] P charger (t)=P cr ·μ·η C (2)

[0017] In the formula, P charger (t) represents the real-time charging power of a single charger; P cr The rated power of the charger; η C For charger efficiency.

[0018] Wind power output equation:

[0019]

[0020] In the formula, P wind v(t) represents the actual electrical power output of the wind turbine at time t; v(t) represents the actual wind speed at time t; v Rv ci and v co These are the rated wind speed, cut-in wind speed, and cut-out wind speed, respectively; P WIND This refers to the rated electrical power of the wind turbine.

[0021] Photovoltaic output power equation:

[0022]

[0023] In the formula, P solar (t) represents the photovoltaic output power; P sr The rated installed capacity of the photovoltaic system; I solar (t) represents the actual solar irradiance on the photovoltaic panel plane at time step t; STC Irradiance under standard test conditions; γ is the power temperature coefficient of the photovoltaic panel; T cell (t) represents the actual operating temperature of the photovoltaic cell at time t; T STC Temperature under standard test conditions; η inv For inverter efficiency;

[0024] Obtaining power from the power grid:

[0025]

[0026] P grid (t)=max(0,P BSS (t)-P solar (t)-P wind (t)) (6)

[0027] In the formula, P BSS (t) represents the total power of the battery swapping station; Total charging power of all chargers; P base Based on standby power; P swap This refers to the operating power of the power swapping equipment; The total operating time of all battery swapping equipment; Δt is the simulation step size; P grid (t) represents the power obtained from the power grid;

[0028] Battery swapping queuing logic:

[0029]

[0030] T wait,i =T depart,i -T arrival,i (9)

[0031] In the formula, The latest battery swap countdown has begun; Q swap Countdown to battery swap begins; E bat Battery state of charge; Estd Standard battery state of charge; L line T is the length of the waiting queue; wait,i T represents the waiting time for the i-th taxi; depart,i Let T be the arrival time of the i-th taxi; arrival,i Let i be the time when the i-th taxi leaves the waiting queue to begin accepting service;

[0032] Specifically, S2 establishes a comprehensive SOC model based on the battery status of the battery swapping station, and proposes a charging scheduling strategy and a dynamic service fee strategy based on the comprehensive SOC status, including the following steps:

[0033] Define the overall SOC status of the battery swapping station. BSS Number of available batteries (SOC≥E) std The ratio of the number of batteries to the total number of batteries is used to model the state of the battery:

[0034]

[0035] In the formula, N a N represents the number of available batteries. b N represents the number of unavailable batteries. bat This represents the total number of batteries;

[0036] The battery state-of-charge update mechanism takes into account the charging efficiency η. c Its update formula is:

[0037]

[0038] In the formula, This represents the updated state of charge of the k-th battery. Its state of charge before the update; P k t represents the charging power allocated to the battery; t represents the charging time. This refers to the battery's rated capacity.

[0039] Comprehensive SOC dependency scheduling strategy:

[0040]

[0041] In the formula, T f X is the transition factor; Z is the SOC threshold parameter; and Z is the transition slope parameter. The planned number of chargers to be activated; N ch y(t) represents the total number of chargers; y(t) represents the hourly scheduling parameters.

[0042] Overall SOC-dependent dynamic service fees:

[0043] P ser =P ser.base +Pser.k ·(1-S BSS (14)

[0044] In the formula, P ser For dynamic service fees; P ser.base Basic service fee; P ser.k SOC sensitivity coefficient;

[0045] In step S3, a joint optimization scheduling model is constructed based on the initial investment cost, annual operating cost, penalty cost, and the parameters and constraints of each device. The constructed joint optimization scheduling model aims to maximize annual net profit and minimize average user waiting time, and establishes a comprehensive multi-objective hierarchical optimization function. Its objective function is as follows:

[0046] F1:Anp=Rev-OC-Pen (15)

[0047]

[0048] In the formula, Anp represents annual net profit (in yuan); OC represents annual operating cost (in yuan); Pen represents penalty amount (in yuan); T avg Average waiting time, in minutes; Total battery swapping time, in minutes; X swap This refers to the number of battery swaps.

[0049] In step S3, the meanings of the variable names are as follows:

[0050] C init =C base +C tr +C grid +C swap +C charger +C bat +C solar +C wind (17)

[0051]

[0052] C swap =c swap ·N swap (20)

[0053] C charger =c charger ·N charger (twenty one)

[0054] C bat =c bat ·N bat (twenty two)

[0055] C solar =csolar ·P solar (twenty three)

[0056] C wind =c wind ·P wind (twenty four)

[0057]

[0058] Rev=P ser ·E T (26)

[0059]

[0060] In the formula, C init The initial investment cost is RMB; C base Infrastructure investment cost, yuan; C tr The cost of the transformer is in yuan; C grid For grid costs, in yuan; C swap The cost of the battery swapping equipment is [amount] yuan; C charger Cost of charging equipment, yuan; C bat Battery equipment cost, yuan; C solar Cost of solar energy, yuan; C wind For wind energy costs, in yuan; c tr The cost per kW is the transformer capacity cost. The maximum power of the battery swapping station is kW; c grid Cost of grid expansion, yuan / kW; c is the maximum power of the power grid, in kW; swap The cost of each battery swapping device is [amount] yuan; N swap c is the number of battery swapping devices. charger The cost per charging device is [amount] yuan; N charger c is the number of charging devices. bat Cost per battery: yuan; N bat c represents the number of batteries. solar The unit cost of photovoltaic power generation is expressed in yuan / (MW·h); P solar Photovoltaic power generation, in MW; c wind The unit cost of wind power generation is expressed in yuan / (MW·h); P wind Wind power generation capacity, MW; OC base Fixed annual operating costs, yuan; k m Maintenance cost ratio factor, P op Unit cost of electricity capacity, yuan / year / kW; E p Electricity purchased from the grid, kWh; T p Real-time electricity price, yuan / kWh; Rev, annual income, yuan; Pser The service fee is dynamic, in yuan / kW; E T Total electricity sales, kW; m is the penalty coefficient, taken as a constant of 1.2; T allow The allowable threshold is set at 15 minutes; DIPP represents the payback period in years.

[0061] In step S4, the self-improved Adaptive Multi-Objective Crayfish Optimization (AMOCO) algorithm is used to solve the joint optimization scheduling model. Relevant data from the battery swapping stations are input, including the objective function of the multi-objective optimization model and related variable constraints, to further derive the configuration and decision variables of the battery swapping station optimization scheduling strategy. With the objectives of maximizing the annual net profit and minimizing the average waiting time of the constructed joint optimization scheduling model for battery swapping stations, the AMOCO algorithm is used to solve the model to obtain the optimal Pareto non-dominated solution, thereby deriving the optimal battery swapping station optimization scheduling strategy. The steps include:

[0062] Chaotic reverse initialization of crayfish population: Set the maximum number of iterations MaxIter, population size N, introduce chaotic initialization to the decision variables to generate an initial population P of N individuals, calculate the reverse individual OX for each individual X in population P, merge P and OP to obtain a temporary population of 2N individuals, evaluate and sort them through step fitness evaluation and non-dominated sorting and external archive set management, select the best N individuals to form the final initial population X(0), each individual represents a battery swapping station optimization scheduling strategy, and use Logistic chaotic mapping to generate chaotic sequence, as shown in the following formula:

[0063] z k+1 =μ·z k (1-z k (29)

[0064] X j (0) = lb j +z j ·(ub j -lb j (30)

[0065] OX i，j =lb j +ub j -X i，j (31)

[0066] In the formula, z k Let X be the value of the k-th chaotic variable; μ is the chaos parameter, which is usually set to 4 to ensure chaos; X j (0) represents the initial value of the j-th dimension decision variable; ub jand lb j Let z represent the upper and lower bounds of the j-th dimension decision variable, respectively; j For the corresponding chaotic variable value; OX i，j X is the reverse solution for the i-th individual in the j-th dimension; i，j This represents the original solution for the i-th individual in the j-th dimension;

[0067] Fitness assessment: The fitness of each crayfish individual was calculated, taking into account constraints. The fitness calculation formula is as follows:

[0068] Fitness (X) i )=[f1(X i ),f2(X i ),…,f k (X i (32)

[0069] In the formula, Fitness(X) i ) refers to the fitness of each individual crayfish; f k (X i ) represents the k-th objective function in the solution X i The value at a given point determines the quality of the solution, which is determined by Pareto dominance.

[0070] Solution X A Dominant solution X B If and only if:

[0071]

[0072] Non-dominated sorting and external archive management: The population is sorted non-dominated according to Pareto dominance and the non-dominated solutions are stored in an external archive. After each iteration, the newly generated non-dominated solutions are compared with the original solutions in the archive, and all non-dominated solutions are retained. If the archive exceeds the capacity limit, the crowding distance of each solution in the archive is calculated and the solution with the smallest crowding distance is removed first.

[0073] Crowding calculation:

[0074]

[0075] In the formula, CD(X i ) represents the congestion distance of solution i; m represents the number of objective functions; and Let be the maximum and minimum values ​​of the k-th objective function in the current non-dominated frontier;

[0076] Temperature determination:

[0077] T = 15·R + 20 (36)

[0078] In the formula, T is the ambient temperature; R is a random number in the range [0,1].

[0079] Search for ways to escape the heat: Execute when T>30 and R<0.5.

[0080]

[0081] In the formula, X shade X represents the location of the cave, which is the current globally optimal solution. G X represents the optimal position obtained through iterations so far. L This is the optimal position for the current population; This represents the individual position in the current iteration; For the individual position in the next iteration; C a For adaptive parameters;

[0082] Competitive search: Execute when T>30 and R≥0.5

[0083]

[0084] z = round(R·(N-1)) + 1 (40)

[0085] In the formula, z represents a randomly selected competitor individual; z represents a randomly selected crayfish individual.

[0086] Foraging search: Execute when T≤30

[0087] X food =X G (41)

[0088]

[0089]

[0090] In the formula, X food The food location during the foraging phase represents the current global optimal solution; Q represents the food size; C3 is the food factor, representing the largest food, with a value of a constant 3; Fitness(X) food ) represents the fitness of the food location; p represents the feeding intensity of the crayfish at the current temperature, with a value between 0 and 1; C1 is the feeding amount adjustment constant, used to scale the range of feeding amount, usually with a value of 0.2; μ is the optimum temperature, with a value of 25; σ is the standard deviation of the temperature distribution, with a value of 3; cos(2πR)-sin(2πR) simulates the random behavior of the crayfish feeding using its second and third legs;

[0091] Levi flight disturbance:

[0092]

[0093] In the formula, a is the step size scaling factor, which is usually taken as 0.01; Levy(β) is the Levy random step size, where β is usually taken as 1.5;

[0094] Improved adaptive parameters:

[0095]

[0096] In the formula, C2 is the original decreasing coefficient; t is the current iteration number;

[0097] Population and Archive Update: The current population X(t) is merged with the offspring population X'(t) generated after the search. The merged population is then subjected to a non-dominated sort, and the external archive is updated. Individuals are selected from the merged population according to the non-dominated sort hierarchy (Front1, Front2, ...) to populate the next generation population X(t+1), until the population size N is reached. For the last selected hierarchy, if the number of individuals exceeds the required number, selections are made from largest to smallest based on the crowding distance to ensure diversity. Finally, a high-quality, highly diverse approximate Pareto optimal solution set is obtained, serving as the strategy set for multi-objective optimal scheduling of the battery swapping station. Attached Figure Description

[0098] Figure 1 This is a flowchart of an optimized scheduling strategy for electric taxi battery swapping stations based on integrated SOC and multiple time scales, as described in this invention.

[0099] Figure 2 This is a diagram of the multi-timescale scheduling framework involved in this invention.

[0100] Figure 3 This is a flowchart of the solution process for the joint optimization scheduling model of the battery swapping stations described in this invention. Specific implementation methods

[0101] With reference to the accompanying drawings of the embodiments of the present invention, the relevant technical solutions will be systematically and comprehensively described below. Figure 1 The overall process of optimized scheduling of battery swapping stations is introduced, and an overall framework for the optimized scheduling of battery swapping stations is built based on the process. Figure 2 It is a multi-time-scale scheduling framework diagram of the battery swapping station system. Based on its time state transition process, the integrated SOC and the two strategies and multi-time scales based on it are modeled. Figure 3This paper introduces the flowchart of the self-improved multi-objective crayfish optimization algorithm, and rationally designs and improves the calculation process to better solve the battery swapping station optimization scheduling problem constructed in this paper. It should be noted that the embodiments represent only some implementation cases of this invention and do not cover all possible technical forms. The specific implementation cases listed in this paper are only for illustrative purposes and should not be considered as limiting definitions of this invention. Based on the disclosed embodiments of this invention, all technical solutions derived by those skilled in the art without creative effort are within the scope of protection of this invention.

[0102] Reference Figure 1 Establish the overall process for optimizing the scheduling of battery swapping stations; based on Figure 2 The content constructs a comprehensive State of Charge (SOC) for battery swapping stations, along with two strategies and a multi-timescale model based on it; according to Figure 3 As shown, the crayfish optimization algorithm is improved into an adaptive multi-objective crayfish optimization algorithm (AMOCO), and its working principle is explained. An optimized scheduling strategy for electric taxi battery swapping stations based on integrated SOC and multiple time scales includes the following steps:

[0103] S1: Modeling the battery charging and discharging model, photovoltaic power generation model, power grid interaction model, and battery swapping service model for the battery swapping station system.

[0104] In this embodiment of the invention, modeling is performed on charging rate, wind and solar power output, grid power, and battery swapping queuing logic.

[0105] S2: Establish a comprehensive SOC model based on the battery status of the battery swapping station, and propose a charging scheduling strategy and a dynamic service fee strategy based on the comprehensive SOC status.

[0106] In this embodiment of the invention, the integrated SOC is modeled based on the number of available batteries, and the state update mechanism of each battery is also taken into account. The charging scheduling strategy and dynamic service fee strategy based on the integrated SOC are transformed into a mathematical model, providing a basis for optimized scheduling.

[0107] S3: In this embodiment of the invention, the constructed joint optimization scheduling model for battery swapping stations is simulated on two time scales: day-ahead global optimization scheduling and intraday rolling optimization scheduling. The objectives are to maximize annual net profit and minimize average waiting time. Constraints include comprehensive SOC state constraints, power balance constraints, battery upper and lower limit constraints, and power upper and lower limit constraints.

[0108] S4: Solving the optimal scheduling problem of battery swapping stations using a self-improved adaptive multi-objective crayfish optimization algorithm. Based on Pareto theory, the original crayfish optimization algorithm is extended to a multi-objective model, designing a basic multi-objective crayfish optimization algorithm framework. In the initialization phase, a chaotic mapping method is introduced to perform chaotic reverse initialization of the population, improving population diversity and global search capability. During the algorithm's iteration process, the parameters in the search are adaptively improved to better balance the algorithm's exploration and development capabilities, avoiding getting trapped in local optima. This series of improvements significantly enhances the algorithm's performance and applicability, providing a more effective solution to the optimal scheduling problem of battery swapping stations.

[0109] Specifically, step S1 includes the following steps:

[0110] S11: Charging rate equation:

[0111]

[0112] In the formula, μ is the charging rate, and E c This refers to the state of charge of the battery during the charging process.

[0113] S12: Real-time charging power of a single charger:

[0114] P charger (t)=P cr ·μ·η C (2)

[0115] In the formula, P charger (t) represents the real-time charging power of a single charger; P cr The rated power of the charger; η C For charger efficiency.

[0116] S13: Wind power output equation:

[0117]

[0118] In the formula, P wind v(t) represents the actual electrical power output of the wind turbine at time t; v(t) represents the actual wind speed at time t; v R v ci and v co These are the rated wind speed, cut-in wind speed, and cut-out wind speed, respectively; P WIND This refers to the rated electrical power of the wind turbine generator.

[0119] S14: Photovoltaic output power equation:

[0120]

[0121] In the formula, P solar(t) represents the photovoltaic output power; P sr The rated installed capacity of the photovoltaic system; I solar (t) represents the actual solar irradiance on the photovoltaic panel plane at time step t; STC Irradiance under standard test conditions; γ is the power temperature coefficient of the photovoltaic panel; T cell (t) represents the actual operating temperature of the photovoltaic cell at time t; T STC Temperature under standard test conditions; η inv This refers to the inverter efficiency.

[0122] S15: Obtain power from the grid:

[0123]

[0124] P grid (t)=max(0,P BSS (t)-P solar (t)-P wind (t)) (6)

[0125] In the formula, P BSS (t) represents the total power of the battery swapping station; Total charging power of all chargers; P base Based on standby power; P swap This refers to the operating power of the power swapping equipment; The total operating time of all battery swapping equipment; Δt is the simulation step size; P grid (t) represents the power obtained from the power grid.

[0126] S16: Battery swapping queuing logic:

[0127]

[0128] T wait,i =T depart,i -T arrival,i (9)

[0129] In the formula, The latest battery swap countdown has begun; Q swap Countdown to battery swapping begins; E bat Battery state of charge; E std Standard battery state of charge; L line T is the length of the waiting queue; wait,i T represents the waiting time for the i-th taxi; depart,i Let T be the arrival time of the i-th taxi; arrival,i Let be the time when the i-th taxi leaves the waiting queue to begin accepting service.

[0130] Specifically, step S2 includes the following steps:

[0131] S21: Integrated SOC Modeling:

[0132] Define the overall SOC status of the battery swapping station. BSS Number of available batteries (SOC≥E) std The ratio of the number of batteries to the total number of batteries is used to model the state of the battery:

[0133]

[0134] In the formula, N a N represents the number of available batteries. b N represents the number of unavailable batteries. bat This represents the total number of batteries;

[0135] S211: Battery state of charge update mechanism:

[0136]

[0137] In the formula, This represents the updated state of charge of the k-th battery. Its state of charge before the update; P k t represents the charging power allocated to the battery; t represents the charging time. This refers to the battery's rated capacity.

[0138] Specifically, step S3 includes the following steps:

[0139] S22: Comprehensive SOC-dependent scheduling strategy:

[0140]

[0141] In the formula, T f X is the transition factor; Z is the SOC threshold parameter; and Z is the transition slope parameter. The planned number of chargers to be activated; N ch y(t) represents the total number of chargers; y(t) represents the hourly scheduling parameters.

[0142] S23: Comprehensive SOC Dependency Dynamic Service Fee:

[0143] P ser =P ser.base +P ser.k ·(1-S BSS (14)

[0144] In the formula, P ser For dynamic service fees; P ser.base Basic service fee; P ser.k SOC sensitivity coefficient;

[0145] Specifically, step S3 includes the following steps:

[0146] S31: Construct a comprehensive multi-objective hierarchical optimization function, where the annual net profit Anp and the average waiting time T are... avg The following formula is used to represent it:

[0147] F1:Anp=Rev-OC-Pen (15)

[0148]

[0149] In the formula, Anp represents annual net profit (in yuan); OC represents annual operating cost (in yuan); Pen represents penalty amount (in yuan); T avg Average waiting time, in minutes; Total battery swapping time, in minutes; X swap This refers to the number of battery swaps.

[0150] S32: The economic cost of scheduling operations is expressed by the following formula:

[0151] C init =C base +C tr +C grid +C swap +C charger +C bat +C solar +C wind (17)

[0152]

[0153] C swap =c swap ·N swap (20)

[0154] C charger =c charger ·N charger (twenty one)

[0155] C bat =c bat ·N bat (twenty two)

[0156] C solar =c solar ·P solar (twenty three)

[0157] C wind =c wind ·P wind (twenty four)

[0158]

[0159] Rev=P ser ·E T (26)

[0160]

[0161] In the formula, C init The initial investment cost is RMB; C base Infrastructure investment cost, yuan; C tr The cost of the transformer is in yuan; C grid For grid costs, in yuan; C swap The cost of the battery swapping equipment is [amount] yuan; C charger Cost of charging equipment, yuan; C bat Battery equipment cost, yuan; C solar Cost of solar energy, yuan; C wind For wind energy costs, in yuan; c tr The cost per kW is the transformer capacity cost. The maximum power of the battery swapping station is kW; c grid Cost of grid expansion, yuan / kW; c is the maximum power of the power grid, in kW; swap The cost of each battery swapping device is [amount] yuan; N swap c is the number of battery swapping devices. charger The cost per charging device is [amount] yuan; N charger c is the number of charging devices. bat Cost per battery: yuan; N bat c represents the number of batteries. solar The unit cost of electricity generated by photovoltaic power generation is expressed in yuan / (MW·h); P solar Photovoltaic power generation, in MW; c wind The unit cost of electricity generated by wind power is expressed in yuan / (MW·h); P wind Wind power generation capacity, MW; OC base Fixed annual operating costs, yuan; k m Maintenance cost ratio factor, P op Unit cost of electricity capacity, yuan / year / kW; E p Electricity purchased from the grid, kWh; T p Real-time electricity price, yuan / kWh; Rev, annual income, yuan; P ser The service fee is dynamic, in yuan / kW; E T Total electricity sales, kW; m is the penalty coefficient, taken as a constant of 1.2; T allow The allowable threshold is set at 15 minutes; DIPP represents the payback period in years.

[0162] S33: System operating constraints:

[0163] S331: Comprehensive SOC state constraints:

[0164] 0≤S BSS ≤1 (29)

[0165] S332: System power balance constraints:

[0166] The total power demand of a battery swapping station consists of charging power, basic operating power, and battery swapping equipment power, and is supplied jointly by renewable energy and the power grid.

[0167]

[0168] In the formula, P BSS This represents the total power of the battery swapping station. For the charging power of all chargers, P base Based on operating power, P represents the power of all power swapping equipment. grid To purchase electricity from the grid, P solar For photovoltaic power generation, P wind This refers to the power output of wind power generation.

[0169] S333: Battery SOC upper and lower limits constraints:

[0170]

[0171] In the formula, and These are the lower and upper limits of SOC, respectively;

[0172] S334: Power Constraints Between Grids

[0173]

[0174] In the formula, The maximum permissible grid interconnection power;

[0175] S335: Charger power and other parameters must not exceed their physical limits.

[0176]

[0177] In the formula, The maximum allowable charging power;

[0178] S336 Renewable Energy Power Constraint:

[0179]

[0180] In the formula, For the maximum permissible photovoltaic power, The maximum permissible wind power;

[0181] Specifically, step S4 includes the following steps:

[0182] S41: Input relevant data about the battery swapping station, including the objective function of the multi-objective optimization model and the relevant variable constraints, and further derive the decision variables for the battery swapping station optimization scheduling strategy.

[0183] S42: Solving Multi-Objective Models

[0184] S421: Chaotic Reverse Initialization of Crayfish Population: Set the maximum number of iterations MaxIter, population size N, introduce chaotic initialization to the decision variables to generate an initial population P with N individuals, calculate the reverse individual OX for each individual X in population P, merge P and OP to obtain a temporary population of 2N individuals, evaluate and sort them through step fitness evaluation and non-dominated sorting and external archive set management, select the best N individuals to form the final initial population X(0), each individual represents a battery swapping station optimization scheduling strategy, and use Logistic chaotic mapping to generate a chaotic sequence, the formula is as follows:

[0185] z k+1 =μ·z k (1-z k (36)

[0186] X j (0) = lb j +z j ·(ub j -lb j (37)

[0187] OX i，j =lb j +ub j -X i，j (38)

[0188] In the formula, z k Let X be the value of the k-th chaotic variable; μ is the chaos parameter, which is usually set to 4 to ensure chaos; X j (0) represents the initial value of the j-th dimension decision variable; ub j and lb j Let z represent the upper and lower bounds of the j-th dimension decision variable, respectively; j For the corresponding chaotic variable value; OX i，j X is the reverse solution for the i-th individual in the j-th dimension; i，j This represents the original solution for the i-th individual in the j-th dimension;

[0189] S422: Fitness Assessment: Calculate the fitness of each individual crayfish, taking constraints into account. The fitness calculation formula is as follows:

[0190] Fitness (X) i )=[f1(X i ),f2(X i ),…,f k (X i (39)

[0191] In the formula, Fitness(X) i ) refers to the fitness of each individual crayfish; f k (X i ) represents the k-th objective function in the solution X i The value at a given point determines the quality of the solution, which is determined by Pareto dominance.

[0192] S423: Non-dominated sorting and external archive management: The population is sorted non-dominated according to the Pareto dominance relationship, and the non-dominated solutions are stored in the external archive. After each iteration, the newly generated non-dominated solutions are compared with the original solutions in the archive, and all non-dominated solutions are retained. If the archive exceeds the capacity limit, the crowding distance of each solution in the archive is calculated, and the solution with the smallest crowding distance is removed first.

[0193] S4231: Congestion Calculation:

[0194]

[0195] In the formula, CD(X i ) represents the congestion distance of solution i; m represents the number of objective functions; and Let be the maximum and minimum values ​​of the k-th objective function in the current non-dominated frontier;

[0196] S424: Rules for Determining Ambient Temperature and Choosing Behavior

[0197] T = 15·R + 20 (41)

[0198] In the formula, T is the ambient temperature; R is a random number in the range [0,1].

[0199] When T>30 and R<0.5, crayfish enter the heat-avoidance stage; when T>30 and R≥0.5, crayfish enter the competition stage; when T≤30, crayfish enter the foraging stage.

[0200] S425: Search for ways to escape the heat:

[0201]

[0202] In the formula, X shade X represents the location of the cave, which is the current globally optimal solution. G X represents the optimal position obtained through iterations so far.L This is the optimal position for the current population; This represents the individual position in the current iteration; For the individual position in the next iteration; C a For adaptive parameters;

[0203] S4251: Improved Adaptive Parameters

[0204]

[0205] In the formula, C2 is the original decreasing coefficient; t is the current iteration number;

[0206] S426: Competitive Search;

[0207]

[0208] z = round(R·(N-1)) + 1 (47)

[0209] In the formula, z represents a randomly selected competitor individual; z represents a randomly selected crayfish individual.

[0210] S427: Foraging search:

[0211] X food =X G (48)

[0212]

[0213]

[0214] In the formula, X food The food location during the foraging phase represents the current global optimal solution; Q represents the food size; C3 is the food factor, representing the largest food, with a value of a constant 3; Fitness(X) food ) represents the fitness of the food location; p represents the feeding intensity of the crayfish at the current temperature, with a value between 0 and 1; C1 is the feeding amount adjustment constant, used to scale the range of feeding amount, usually with a value of 0.2; μ is the optimum temperature, with a value of 25; σ is the standard deviation of the temperature distribution, with a value of 3; cos(2πR)-sin(2πR) simulates the random behavior of the crayfish feeding using its second and third legs;

[0215] S4271: Levy flight disturbance:

[0216]

[0217] In the formula, a is the step size scaling factor, which is usually taken as 0.01; Levy(β) is the Levy random step size, where β is usually taken as 1.5;

[0218] S428: Population Update and Archive Update:

[0219] S4281: Merging populations:

[0220] X combined =X(t)∪X′(t) (54)

[0221] S4282: Non-dominated sorting: Calculate the dominated count (the number of solutions that dominate it) for each solution. The solution with a dominated count of 0 is the first front (Front1), and subsequent fronts (Front2, Front3, ...) are generated iteratively.

[0222] S4283: Fill by Front Level: Prioritize adding Front1 to the next generation X(t+1) until the population size N is reached. For the last selected level, if the number of individuals exceeds the requirement, select from largest to smallest based on crowding distance to ensure diversity.

[0223] S4284: Update external archive:

[0224] Archive=UpdateNonDominated(X(t)∪X′(t)) (55)

[0225] Add all solutions of the current population X to the archive, perform non-dominated sorting on the merged solution set, remove dominated and duplicate solutions, and retain only all non-dominated solutions (i.e., the first frontier).

[0226] S4285: Archive Removal:

[0227] If |Archive|>N max :PruneByCrowding(Archive) (56)

[0228] The solutions in the archive are sorted by the numerical value of each objective function, and the crowding distance of each solution is calculated. The crowding distance of the boundary solutions (i.e., the solutions corresponding to the minimum and maximum values ​​of each objective) is set to infinity to ensure that these boundary solutions are forcibly preserved. Then, all solutions are sorted in descending order of crowding distance, and only the top N solutions with the largest crowding distances are retained.

[0229] S429: Finally, a high-quality, highly diverse set of solutions for approximate Pareto optimality is obtained, from which users can select appropriate solutions based on their actual needs.

[0230] It should be understood by those skilled in the art that this invention is not limited to the content described in the foregoing specific embodiments. Without departing from the essential spirit or basic characteristics of this invention, it can also be implemented in other specific forms. The given embodiments are merely illustrative and do not constitute a limitation on the scope of protection of this invention. The scope of protection of this invention is defined by the appended claims and their legal equivalents. Any modifications made under the guidance of the inventive concept are covered by the claims. The reference numerals in the claims do not affect the scope of interpretation of the claims. Although this specification describes various solutions with specific embodiments, those skilled in the art can combine the technical solutions in different embodiments to form other achievable implementations.

Claims

1. An optimized scheduling strategy for electric taxi battery swapping stations based on integrated SOC and multiple time scales, characterized in that, Includes the following steps: S1: Modeling the battery charging and discharging model, photovoltaic power generation model, power grid interaction model, and battery swapping service model for the battery swapping station system; S2: Establish a comprehensive SOC model based on the battery status of the battery swapping station, and propose a charging scheduling strategy and a dynamic service fee strategy based on the comprehensive SOC status. S3: Construct a joint optimization scheduling model for battery swapping stations that takes into account both comprehensive SOC optimization management and day-ahead and intraday multi-timescale scheduling efficiency; S4: A self-improved Adaptive Multi-Objective Crayfish Optimization (AMOCO) algorithm is used for optimized scheduling of battery swapping stations. Based on Pareto theory, the original crayfish optimization algorithm is extended to a multi-objective model, resulting in a basic framework. In the initialization phase, a Logistic chaotic mapping method is introduced to perform chaotic inverse initialization of the population, and parameters are adaptively improved during the heat avoidance search process.

2. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 1, characterized in that: The battery swapping station system modeling includes battery charging model, photovoltaic power generation model, power grid interaction model, and battery swapping service model, and the modeling and analysis of these four types of processes are carried out respectively.

3. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 1, characterized in that: The integrated SOC model, as well as the charging scheduling strategy and dynamic service fee strategy based on the integrated SOC state, respectively model the integrated SOC state and the two strategies.

4. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 2, characterized in that: The construction of the battery charging model, photovoltaic power generation model, grid interaction model, and battery swapping service model includes the following steps: Charging rate equation: In the formula, μ is the charging rate, and E c This refers to the state of charge of the battery during the charging process. Real-time charging power of a single charger: P charger (t)=P cr ×m×h C (2) In the formula, P charger (t) represents the real-time charging power of a single charger; P cr The rated power of the charger; h C For charger efficiency. Wind power output equation: In the formula, P wind v(t) represents the actual electrical power output of the wind turbine at time t; v(t) represents the actual wind speed at time t; v R v ci and v co These are the rated wind speed, cut-in wind speed, and cut-out wind speed, respectively; P WIND This refers to the rated electrical power of the wind turbine generator. Photovoltaic output power equation: In the formula, P solar (t) represents the photovoltaic output power; P sr The rated installed capacity of the photovoltaic system; I solar (t) represents the actual solar irradiance on the photovoltaic panel plane at time step t; STC Irradiance under standard test conditions; γ is the power temperature coefficient of the photovoltaic panel; T cell (t) represents the actual operating temperature of the photovoltaic cell at time t; T STC Temperature under standard test conditions; h inv For inverter efficiency; Obtaining power from the power grid: P grid (t)=max(0,P BSS (t)-P solar (t)-P wind (t)) (6) In the formula, P BSS (t) represents the total power of the battery swapping station; Total charging power of all chargers; P base Based on standby power; P swap This refers to the operating power of the power swapping equipment; The total operating time of all battery swapping equipment; Δt is the simulation step size; P grid (t) represents the power obtained from the power grid; Battery swapping queuing logic: T wait,i =T depart,i -T arrival,i (9) In the formula, The latest battery swap countdown has begun; Q swap Countdown to battery swapping begins; E bat Battery state of charge; E std Standard battery state of charge; L line T is the length of the waiting queue; wait,i T represents the waiting time for the i-th taxi; depart,i Let T be the arrival time of the i-th taxi; arrival,i Let be the time when the i-th taxi leaves the waiting queue to begin accepting service.

5. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 3, characterized in that: The establishment of the integrated SOC model and the charging scheduling strategy and dynamic service fee strategy based on the integrated SOC state include the following details: Define the overall SOC status of the battery swapping station. BSS Number of available batteries (SOC 3E) std The ratio of the number of batteries to the total number of batteries is used to model the state of the battery: In the formula, N a N represents the number of available batteries. b N represents the number of unavailable batteries. bat This represents the total number of batteries; The battery state-of-charge update mechanism takes into account the charging efficiency h c Its update formula is: In the formula, This represents the updated state of charge of the k-th battery. Its state of charge before the update; P k t represents the charging power allocated to the battery; t represents the charging time. This refers to the battery's rated capacity. Comprehensive SOC dependency scheduling strategy: In the formula, T f X is the transition factor; Z is the SOC threshold parameter; and Z is the transition slope parameter. The planned number of chargers to be activated; N ch y(t) represents the total number of chargers; y(t) represents the hourly scheduling parameters. Overall SOC-dependent dynamic service fees: P ser =P ser.base +P ser.k ×(1-S BSS ) (14) In the formula, P ser For dynamic service fees; P ser.base Basic service fee; P ser.k This is the SOC sensitivity coefficient.

6. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 1, characterized in that: The current global optimization scheduling and intraday rolling optimization scheduling are two different time scales. The day-ahead global optimization scheduling layer employs a multi-objective crayfish optimization algorithm, with a 24-hour cycle and a 1-hour step size, to simultaneously optimize the hardware configuration and scheduling strategy of the battery swapping stations. Decision variables include four hardware investment parameters: number of batteries, number of chargers, photovoltaic capacity, and number of battery swapping devices, as well as 24 hourly charging intensity parameters, a SOC threshold, and a transition slope parameter. This layer seeks the optimal trade-off between maximizing annual net profit and minimizing average waiting time by maintaining a Pareto solution set and calculating congestion distance. Its output is a set of non-dominated solutions, each containing a complete equipment configuration scheme and corresponding all-day scheduling strategy parameters, providing an optimized global operating framework for the next layer. Intraday Rolling Optimization Scheduling: The intraday layer performs high-precision simulations based on the day-ahead strategy, simulating actual operation in 15-minute increments. This layer receives the configuration and strategy parameters from the day-ahead solution, dynamically determines the number of chargers to start and stop by generating random demand sequences, microscopically simulating battery swapping service queuing, and calculating the system's SOC status in real time, while accurately recording waiting time and power consumption. The simulation strictly adheres to battery charging characteristics and power balance principles, evaluating the real performance of each solution under uncertainty. If a solution violates the waiting time constraint, it is marked as infeasible. Finally, precise indicators such as average annual profit and average waiting time are fed back, forming an "optimization-simulation-evaluation" closed loop to ensure the practicality and robustness of the day-ahead strategy.

7. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 1, characterized in that: A joint optimization scheduling model is constructed based on the initial investment cost, annual operating cost, penalty cost, and parameters and constraints of each device. The constructed joint optimization scheduling model aims to maximize annual net profit and minimize average user waiting time, and establishes a comprehensive multi-objective hierarchical optimization function. The objective function is as follows: F1:Anp=Rev-OC-Pen (15) In the formula, Anp represents annual net profit (in yuan); OC represents annual operating cost (in yuan); Pen represents penalty amount (in yuan); T avg Average waiting time, in minutes; Total battery swapping time, in minutes; X swap This refers to the number of battery swaps. The economic cost of scheduling operations is expressed by the following formula: C init =C base +C tr +C grid +C swap +C charger +C bat +C solar +C wind (17) C swap =c swap ×N swap (20) C charger =c charger ×N charger (21) C bat =c bat ×N bat (22) C solar =c solar ×P solar (23) C wind =c wind ×P wind (24) Rev=P ser ×E T (26) In the formula, C init The initial investment cost is RMB; C base Infrastructure investment cost, yuan; C tr The cost of the transformer is in yuan; C grid For grid costs, in yuan; C swap The cost of the battery swapping equipment is [amount] yuan; C charger Cost of charging equipment, yuan; C bat Battery equipment cost, yuan; C solar Cost of solar energy, yuan; C wind For wind energy costs, in yuan; c tr The cost per kW is the transformer capacity cost. The maximum power of the battery swapping station is kW; c grid Cost of grid expansion, yuan / kW; c is the maximum power of the power grid, in kW; swap The cost of each battery swapping device is [amount] yuan; N swap c is the number of battery swapping devices. charger The cost per charging device is [amount] yuan; N charger c is the number of charging devices. bat Cost per battery: yuan; N bat c represents the number of batteries. solar The unit cost of photovoltaic power generation is expressed in yuan / (MW×h); P solar Photovoltaic power generation, in MW; c wind The unit cost of wind power generation is expressed in yuan / (MW×h); P wind Wind power generation capacity, MW; OC base Fixed annual operating costs, yuan; k m Maintenance cost ratio factor, P op Unit cost of electricity capacity, yuan / year / kW; E p Electricity purchased from the grid, kWh; T p Real-time electricity price, yuan / kWh; Rev, annual income, yuan; P ser The service fee is dynamic, in yuan / kW; E T Total electricity sales, kW; m is the penalty coefficient, taken as a constant of 1.2; T allow The allowable threshold is set to 15 minutes; DIPP is the payback period in years.

8. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 1, characterized in that: The optimization model in step S3 must satisfy the following system operation constraints: Comprehensive SOC state constraints: 0≤S BSS ≤1 (29) System power balance constraints: The total power demand of a battery swapping station consists of charging power, basic operating power, and battery swapping equipment power, and is supplied jointly by renewable energy and the power grid. In the formula, P BSS This represents the total power of the battery swapping station. For the charging power of all chargers, P base Based on operating power, P represents the power of all power swapping equipment. grid To purchase electricity from the grid, P solar For photovoltaic power generation, P wind This refers to the power output of wind power generation. Battery SOC upper and lower limits constraints: In the formula, and These are the lower and upper limits of SOC, respectively; Power constraints: In the formula, The maximum permissible grid interconnection power; The maximum allowable charging power; The maximum permissible photovoltaic power; The maximum permissible wind power output.

9. The electric taxi battery swapping station optimization scheduling strategy based on integrated SOC and multiple time scales as described in claim 1, characterized in that: The improved adaptive multi-objective crayfish optimization algorithm includes several aspects: initializing the crayfish population, fitness assessment, temperature determination, heat avoidance and competition search, foraging search, and population and archive updates. The specific steps are as follows: Chaotic reverse initialization of crayfish population: Set the maximum number of iterations MaxIter, population size N, introduce chaotic initialization to the decision variables to generate an initial population P of N individuals, calculate the reverse individual OX for each individual X in population P, merge P and OP to obtain a temporary population of 2N individuals, evaluate and sort them through step fitness evaluation and non-dominated sorting and external archive set management, select the best N individuals to form the final initial population X(0), each individual represents a battery swapping station optimization scheduling strategy, and use Logistic chaotic mapping to generate chaotic sequence, as shown in the following formula: With k+1 =m×z k (1-z k ) (36) X j (0)=lb j +z j ×(ub j -lb j ) (37) OX i，j =lb j +ub j -X i，j (38) In the formula, z k Let X be the value of the k-th chaotic variable; μ is the chaos parameter, which is usually set to 4 to ensure chaos; X j (0) represents the initial value of the j-th dimension decision variable; ub j and lb j Let z represent the upper and lower bounds of the j-th dimension decision variable, respectively; j For the corresponding chaotic variable value; OX i，j X is the reverse solution for the i-th individual in the j-th dimension; i，j This represents the original solution for the i-th individual in the j-th dimension; Fitness assessment: The fitness of each crayfish individual was calculated, taking into account constraints. The fitness calculation formula is as follows: Fitness(X i )=[f1(X i ),f2(X i ),…,f k (X i )] (39) In the formula, Fitness(X) i ) refers to the fitness of each individual crayfish; f k (X i ) represents the k-th objective function in the solution X i The value at a given point determines the quality of the solution, which is determined by Pareto dominance. Solution X A Dominant solution X B If and only if: Non-dominated sorting and external archive management: The population is sorted non-dominated according to the Pareto dominance relationship, and the non-dominated solutions are stored in the external archive. After each iteration, the newly generated non-dominated solutions are compared with the original solutions in the archive, and all non-dominated solutions are retained. If the archive exceeds the capacity limit, the crowding distance of each solution in the archive is calculated, and the solution with the smallest crowding distance is removed first. Crowding calculation: In the formula, CD(X i ) represents the congestion distance of solution i; m represents the number of objective functions; and Let be the maximum and minimum values ​​of the k-th objective function in the current non-dominated frontier; Temperature determination: T = 15 × R + 20 (43) In the formula, T is the ambient temperature; R is a random number in the range [0,1]. Search for ways to escape the heat: Execute when T>30 and R<0.

5. In the formula, X shade X represents the location of the cave, which is the current globally optimal solution. G X represents the optimal position obtained through iterations so far. L This is the optimal position for the current population; This represents the individual position in the current iteration; For the individual position in the next iteration; C a For adaptive parameters; Competitive search: Execute when T > 30 and R ≥ 30.5 z = round(R × (N-1)) + 1 (47) In the formula, z represents a randomly selected competitor individual; z represents a randomly selected crayfish individual. Foraging search: Execute when T≤30 X food =X G (48) In the formula, X food The food location during the foraging phase represents the current global optimal solution; Q represents the food size; C3 is the food factor, representing the largest food, with a value of a constant 3; Fitness(X) food ) represents the fitness of the food location; p represents the feeding intensity of the crayfish at the current temperature, with a value between 0 and 1; C1 is the feeding amount adjustment constant, used to scale the range of feeding amount, usually with a value of 0.2; μ is the optimum temperature, with a value of 25; s is the standard deviation of the temperature distribution, with a value of 3; cos(2πR)-sin(2πR) simulates the random behavior of the crayfish feeding using its second and third legs; Levi flight disturbance: In the formula, a is the step size scaling factor, which is usually taken as 0.01; Levy(β) is the Levy random step size, where β is usually taken as 1.5; Improved adaptive parameters: In the formula, C2 is the original decreasing coefficient; t is the current iteration number; Population and Archive Update: The current population X(t) is merged with the offspring population X'(t) generated after the search. The merged population is then subjected to a non-dominated sort, and the external archive is updated. Individuals are selected from the merged population according to the non-dominated sort hierarchy (Front1, Front2, ...) to populate the next generation population X(t+1), until the population size N is reached. For the last selected hierarchy, if the number of individuals exceeds the required number, selections are made from largest to smallest based on the crowding distance to ensure diversity. Finally, a high-quality, highly diverse approximate Pareto optimal solution set is obtained, serving as the strategy set for multi-objective optimal scheduling of the battery swapping station.

Images